Champagne Harvest 2026: Less Wine, Better Acidity

On 6 August, Vincent Finot stood in his Montgueux vines with secateurs in hand. Nobody in Champagne had ever picked this early. He had to request a derogation to start before the official date, and he got it, because his measurements showed the acidity falling away. No grower wants to see that.

Six days later the Comité Champagne opened the champagne harvest 2026 officially. 12 August. The earliest since records began, somewhere early in the last century.

Then came the headlines. Reuters, CNBC, the rest: heat, drought, frost, and a region losing its freshness. Factually all of it holds up. Two numbers from this vintage point the other way, though.

The ceiling nobody reaches

On 22 July, Champagne set the marketable yield for 2026 at 8,800 kilos per hectare. Roughly 250 million bottles. That is the fourth cut in a row: 12,000 in 2022, 11,400 in 2023, 10,000 in 2024, 9,000 in 2025. The lowest level since the covid-flattened 2020.

That is how it was reported, too. Region cuts quota, region manages stock, region protects price.

Except a quota is a ceiling, not a floor. And according to Sébastien Dubuisson of the Comité Champagne itself, actual yields this year will land around 7,000 kilos per hectare. Well under the 8,800 officially permitted. The 250 million bottles will not materialise, and not because of an administrative decision. There simply wasn’t that much hanging on the vines.

How much the frost actually cost depends on who you ask and what is being counted. The Comité itself put it at 38 percent of buds destroyed. David Chatillon of the Union des Maisons de Champagne later cited 43 percent of production potential, a different number over a different denominator. Reuters reported roughly 30 percent of the crop, on the Comité’s figures. Three numbers that do not contradict each other and do not measure the same thing either. Heat and water stress did the rest. Thibaud Brocard of Pierre Brocard put it plainly to Tom Hewson: compared to a normal vintage, there is almost no liquid inside the berries. Small fruit with thick skins, and little juice underneath.

This is the difference between “the region chose to make less” and “the region could not make more”. The first is inventory policy. The second is a vintage.

The acidity that held

Now the figure that genuinely surprised me.

The Réseau Matu, the Comité’s ripeness monitoring network, opened its readings at an average of 8.2 percent potential alcohol. A week later the appellation average sat at 10.2 percent potential alcohol against a total acidity of 9 g/l. That works out to a sugar/acid ratio of 19.7, just inside the limit of 20 that marks the ideal picking window.

Dubuisson reported that acidity not only survived, but at the same alcohol level came in slightly higher than last year.

Set that against the international coverage and a gap opens up. The thrust there was that heat is stripping the acidity and with it Champagne’s identity, and that identity does rest on tension: high acid, low sugar, finesse. But the Comité’s own measurement series says the tension was there this year, so far.

The rush comes down to sugar climbing faster than acidity falls. The Réseau Matu warned that picking from around 17 August would push potential alcohol to 12.5 percent or beyond. Its advice was to leave the ripest parcels hanging and start with the plots showing lower degrees and higher acidity. Not picking because the fruit is ripe, but picking before it gets too ripe. That is a different craft than it was fifteen years ago.

Where averages mislead

A warning belongs here, and it applies to my own argument above as much as anyone’s.

That 9 g/l is an appellation average. Finot in Montgueux picked at 11 percent potential alcohol with a total acidity of 4 g/l. His acidity collapsed after 30 millimetres of rain fell. One parcel, one Monday, and the average tells you nothing useful.

Look at the spread and it gets worse. Where frost was measured regionally, the Aisne came in at 65 to 85 percent bud loss against 38 percent for the region as a whole. In a year with spring frost, summer heat and highly local rainfall, an average mostly tells you that extremes exist. Anyone calling 2026 a great vintage right now is reading a spreadsheet, not tasting wine.

Then there is that 7,000 kilos per hectare. It is a mid-harvest projection, not a final count. I take it seriously because it comes from the Comité itself, and I also note that Champagne has no interest whatsoever in understating scarcity. A region working down its stocks tells the scarcity story willingly. Settle up in spring.

What actually lands in your bottle

Barely any of it, and almost no harvest report covers that.

Most Champagne you buy is non-vintage: a blend across years, with reserve wines absorbing the shocks. Exactly what a house reaches for in a year like this. You will not taste 2026 as 2026, unless you open a millésime five years from now, and that only applies if there is enough worth declaring.

One decision does land closer to your glass. An arrêté dated 7 August, published in the Journal officiel on 12 August, lifts the ceiling on bought-in grapes, must and wine to 15 percent for this campaign. The usual ceiling is 5 percent. It is the relief valve for a small harvest, entirely legal, and it also means bottles from 2026 will more often contain something that came from someone other than the name on the label. For the large houses this changes nothing about how they already work. For small récoltants, whose entire proposition is that the grower and the maker are the same person, it sits less comfortably. What that order does and does not govern is a story of its own. I will come back to it.

And the price

It stays flat.

That reads as a contradiction against a harvest smaller than planned, until you put the second movement beside it. Volumes are falling too. Global shipments in the first half of 2026 reached 107.1 million bottles, up 1.2 percent year on year, and that growth came from exports rather than domestic demand.

A shrinking harvest against shrinking demand produces a flat price. A flat price at a high level, though. Real Champagne starts around 15 euro on a Dutch supermarket shelf, but a bottle you would actually want to pour sits closer to 40.

If you want the same method for less money, the value sits outside Champagne. Barth’s Riesling Brut from the Rheingau does bottle fermentation with 24 months on the lees for 19.95 euro, and it is the sharpest buy I found in this bracket this year. Crémant d’Alsace from François Schwach runs around 18 to 20 euro, though the Blanc de Blancs was sold out at the Dutch retailer on 11 August.

What 2026 does to the style over the longer term is a separate question. I wrote about that when Champagne started creeping towards 14 percent alcohol.

For now: agroclimatologist Serge Zaka calculated that French harvest dates have moved forward some 21 days over fifty years. Denis Velut, 67 and a lifetime in the region, remembers harvests running into late September and October.

August is no longer the exception, and that made none of the headlines.

Sources

Primary source: The 2026 Champagne Harvest, Tom Hewson (Six Atmospheres) — 20 August 2026
Additional: Champagne Harvest Earliest Ever, Wine-Searcher — 13 August 2026
Additional: Champagne’s Readiness for Climate Change, Terroir Champagne — 11 August 2026
Additional: Record-early champagne harvest highlights climate change impact, Reuters — 17 August 2026
Additional: Extreme heat spurs earliest-ever Champagne harvest, CNBC — 15 August 2026
Additional: Champagne caps 2026 harvest at its lowest yield since the pandemic, Winecap — 5 August 2026
Additional: Champagne will start its 2026 grape harvest earlier than ever recorded, Vinetur — 14 August 2026
Additional: Un rendement attendu de 7.000 kilos par hectare en Champagne, L’Union — 12 August 2026 (38 percent bud loss, regional spread)
Additional: Frost destroyed 43% of potential production, Vinetur — 20 August 2026
Additional: Arrêté du 7 août 2026 (15 percent purchase ceiling), Journal officiel — published 12 August 2026